Sir

In his Millennium Essay (Nature 408, 293; 2000), Jim Smith proposes that ecological theory is generally untestable, and therefore that ecology should concentrate less on theoretical explanation and more on finding applied solutions to humankind's environmental problems.

Although his concern for solving environmental problems is widely shared by theoretical and empirical ecologists alike, Smith's call for use of Robert Peters's 'predictive ecology' in place of a more conceptually grounded approach runs the risk of leading ecology into a dead end of blind empiricism.

Smith overstates Karl Popper's influence over the philosophy of science. Science is a much broader and less formally rigorous enterprise than Popper envisioned. In The Structure of Scientific Revolutions (University of Chicago Press, Chicago, 1970) Thomas Kuhn, for example, argued that a large component of cultural subjectivity determines the outcome of the scientific enterprise. His concept of paradigm shifts implicitly assumes that political interactions among scientists are at least as important as empirical verification of hypotheses in deciding the outcome of scientific progress. Science can still uncover much about the natural world, even if it is such a culturally dominated social system. Neither Popper's views nor a less restrictive philosophy of science justifies the abandonment of conceptual work in ecology.

Progress in ecology has been hampered in the past by assuming that complex ecological systems can be adequately described by relatively simple linearized models, but even those who originated this approach realized its limited usefulness. Significant advances in science are not based solely on the success of a model in predicting some unknown phenomenon.

Smith's example of relativity is a case in point. The real significance of relativity to physics was not its ability to predict the bending of light by large, massive celestial bodies. Rather, relativity revolutionized physics by revealing an underlying conceptual unity to disparate phenomena. Relativity did not supplant classical physics, it simply limited the spatial and temporal scales to which the rich theoretical framework of classical physics could be applied with acceptable precision.

Ecology will make real scientific progress only by developing a rigorous conceptual framework that can be tested with appropriately sophisticated statistical methodology. Ecological systems are indeed complex. Simplistic semi-empirical relationships based on vague, poorly parameterized, linear statistical models of single systems provide only approximate or temporary solutions to environmental problems. They do not determine which problems are of greatest importance or how limited resources might be optimally allocated.

Only a larger, more comprehensive conceptual framework can provide such guidance. Explorations of ecological theory are nice work and are essential to the progress of ecology as science.